Camel中Splitter和Aggregator的使用

最近公司要用camel来切分字符串,然后聚合起来,判断是否处理完成。容器是用的Servicemix。

blueprint.xml

<?xml version="1.0" encoding="UTF-8"?>
<blueprint xmlns="http://www.osgi.org/xmlns/blueprint/v1.0.0" 
           xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
           xsi:schemaLocation="http://www.osgi.org/xmlns/blueprint/v1.0.0 http://www.osgi.org/xmlns/blueprint/v1.0.0/blueprint.xsd" xmlns:ns2="null" xmlns:ns3="null" xmlns:ns4="null" xmlns:ns5="null" xmlns:ns6="null" xmlns:ns7="null">
    
    <bean id="Aggregated" class="com.bosame.TicketCenter.Application.OrderProcedure.Aggregated">
    </bean>

    <bean id="aggregatorStrategy" class="com.bosame.TicketCenter.Application.OrderProcedure.StringAggregationStrategy"/>

    <bean id="OrderSplitBean" class="com.bosame.TicketCenter.Application.OrderProcedure.OrderSplit"></bean>

    <camelContext id="OrderCamel"   xmlns="http://camel.apache.org/schema/blueprint" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
                    xsi:schemaLocation="http://camel.apache.org/schema/blueprint http://camel.apache.org/schema/blueprint/camel-blueprint.xsd">
        <route>
            <from uri="vm:ReceiveBatchOrder"/>
            <setHeader headerName="gid">
                <simple>${date:now:HH_mm_ss}</simple>
            </setHeader>
            <split>
                <method bean="OrderSplitBean" method="splitOrders"/>
                <setHeader headerName="gid">
                    <simple>${in.header.gid}</simple>
                </setHeader>
                <to uri="vm:ReceiveOrder"/>
            </split>
        </route>
        <route>
            <from uri="vm:ReceiveOrder"/>
            <to uri="bean:OrderVerifierBean" />
            <aggregate strategyRef="aggregatorStrategy" eagerCheckCompletion="true" completionTimeout="3000">
                <correlationExpression>
                    <header>gid</header>
                </correlationExpression>
                <completionSize>
                    <header>CamelSplitSize</header>
                </completionSize>
                <to uri="bean:Aggregated"/>
            </aggregate>
        </route>
    </camelContext>
</blueprint>

 

这里切分的时候是使用的一个bean对象来切分的,也可以使用正则表达式或者字符串。

就是把"<method bean="OrderSplitBean" method="splitOrders"/>"这个改成"<tokenize token="@"/>",切分的时候就会以"@"来切分。

加入<setHeader headerName="gid">
                <simple>${date:now:HH_mm_ss}</simple>
        </setHeader>
这行代码是为了方便聚合的时候使用,因为camel聚合的时候是根据头部相同的来聚合的。

<correlationExpression>
                    <header>gid</header>
</correlationExpression>
这个是表示聚合条件,以头部相同的来聚合。

<completionSize>
                    <header>CamelSplitSize</header>
</completionSize>
这个是聚合完成条件。以切分的大小为完成条件。CamelSplitSize这个参数在camel的2.9以后已经加入到了camel的头部header 里面。详细请参考链接http://camel.apache.org/splitter.html

completionTimeout="3000"这个参数是超时时间,当3秒钟还没聚合完时,就算超时。
strategyRef="aggregatorStrategy" 这个是聚合的策略,要自己编写一个类继承AggregationStrategy的。

里面就写要怎样聚合,我这里是简单把2个值相加。

<to uri="bean:Aggregated"/>这个是最后聚合完成了,要跳转的bean。可以自己随便写。

OrderSplit.java

package com.TicketCenter.Application.OrderProcedure;

import java.util.List;

/**
 * 切分订单
 * @author lym
 */
public class OrderSplit {
    public List<String> splitOrders(List<String> orderStrList) {
        return orderStrList;
    }
}

StringAggregationStrategy.java

package com.TicketCenter.Application.OrderProcedure;

import org.apache.camel.Exchange;
import org.apache.camel.processor.aggregate.AggregationStrategy;

/**
 *
 * @author Administrator
 */
public class StringAggregationStrategy implements AggregationStrategy {

    @Override
    public Exchange aggregate(Exchange oldExchange, Exchange newExchange) {
        if (oldExchange == null) {
            return newExchange;
        }
        oldExchange.getIn().setBody(newExchange.getIn().getBody(Order.class));
        return oldExchange;
    }
}

 

你可能感兴趣的:(osgi,camel,servicemix)